Sulfur isotope measurement of sulfate and sulfide by high-resolution MC-ICP-MS

We have developed a technique for the accurate and precise determination of (34)S/(32)S isotope ratios (delta(34)S) in sulfur-bearing minerals using solution and laser ablation multiple-col lector inductively coupled plasma mass spectrometry (MC-ICP-MS). We have examined and determined rigorous corrections for analytical difficulties such as instrumental mass bias, unresolved isobaric interferences, blanks, and laser ablation- and matrix-induced isotopic fractionation. Use of high resolution sector-held mass spectrometry removes major isobaric interferences from O(2)(+). Standard-sample bracketing is used to correct for the instrumental mass bias of unknown samples. Background on sulfur masses arising from memory effects and residual oxygen-tailing are typically minor (< 0.2 parts per thousand, within analytical error), and are mathematically removed by on-peak zero subtraction and by bracketing of samples with standards determined at the same signal intensity (within 20%). Matrix effects are significant (up to 0.7 parts per thousand) for matrix compositions relevant to many natural sulfur-bearing minerals. For solution analysis, sulfur isotope compositions are best determined using purified (matrix-clean) sulfur standards and sample solutions using the chemical purification protocol we present, For in situ analysis, where the complex matrix cannot be removed prior to analysis, appropriately matrix-matching standards and samples removes matrix artifacts and yields Sulfur isotope ratios consistent with conventional techniques using matrix-clean analytes. Our method enables solid samples to be calibrated against aqueous standards; a consideration that is important when certified, isotopically-homogeneous and appropriately matrix-matched solid standards do not exist. Further, bulk and in situ analyses can be performed interchangeably in a single analytical session because the instrumental setup is identical for both. We validated the robustness of our analytical method through multiple isotope analyses of a range of reference materials and have compared these with isotope ratios determined using independent techniques. Long-term reproducibility of S isotope compositions is typically 0.20%. and 0.45%. (2 sigma) for solution and laser analysis, respectively. Our method affords the opportunity to make accurate and relatively precise S isotope measurement for a wide range of sulfur-bearing materials, and is particularly appropriate for geologic samples with complex matrix and for which high-resolution in situ analysis is critical. (C) 2008 Elsevier B.V. All rights reserved.